A System for Printed Persian Documents

Introducing recognition systems and retrieval systems for Farsi printed document images and categorizing conducted researches with identifying strengths and weaknesses points of each category and presenting a retrieval system for Farsi printed document images in a new way.

This paper is an applied research.An application designed and implied for Farsi printed document images retrieval. A new method in segmentation approach presented and implemented. A database including 50 Farsi documents scanned in 5 fonts provided for training and testing levels. Half of this database is used for training and other half for testing.

Persian printed document image recognition and retrieval systems follow one of three approaches: based on segmentation, based on sub-word shape and hybrid approach. The first approachhas received less attention dueto specific challenges. In this paper a system based on segmentation proposed and implemented and the results presented.

Results show that proposed system is anuseful method to Implement recognition systems and retrieval systems for Farsi printed document images.